The hypothesis to be tested in this project is that there are tumor specific antigenic determinants recognized, in association with MHC, by autologous tumor-specific T-cell clones. The overall objective is to identify these tumor-specific antigenic determinants. The specific aims are: 1) to clone and determine the sequence of the "tumor-specific antigen" involved: a) to construct on a retroviral vector a cDNA library from an appropriate melanoma tumor cell clone expressing "tumor-specific antigen"; b) to infect autologous recipient cells (EBV B-cell lines or melanoma clones not expressing the "tumor-specific antigen") with the retrovirus; c) to identify recipient cells expressing "tumor-specific antigen" by their ability to stimulate autologous tumor-specific TIL T-cell clones. To confirm the expression of "tumor-specific antigen" by the recipient cells by their susceptibility to lysis by autologous tumor-specific TIL T cell clones; to clone the gene encoding for the "tumor-specific antigen" from the recipient cells and determine its sequence; e) to employ the cosmid library/P815.P1.HTR transfection procedure of Boon et al as an alternative approach, in the event that the retroviral cDNA library does not permit the expression of "tumor-specific antigen" on recipient cells. 2) To identify the peptide of the "tumor-specific antigen" that is recognized in association with self-MHC by autologous tumor-specific T cells. These studies are hoped to improve on understanding of the response of the immune system to tumor cells, the immunogenicity of tumors, and the mechanism(s) of immune surveillance and tumor containment.